Sunk Costs and the Disposition Effect

You bought a stock at $50. It’s now $30. Every fibre of your being wants to wait “until it gets back to what I paid.” That sentence — so reasonable, so common — is one of the most expensive mistakes in investing, and it’s built out of two psychological traps stacked on top of each other. The first is the sunk-cost fallacy: letting money you can’t get back steer a decision it has no business touching. The second is the disposition effect: the strange, well-documented habit of cashing in your winners early while clinging to your losers for dear life. This lesson takes both apart, shows you the data, and hands you one clean test that defuses them.

Sunk costs — money you can’t get back shouldn’t get a vote

A sunk cost is money (or time, or effort) you have already spent and cannot recover, no matter what you do next. The defining feature is that it’s the same whether you continue or quit — which means, logically, it should have zero influence on your next choice. Only future costs and benefits can differ between your options, so only future costs and benefits should drive the decision. Yet humans treat sunk costs as if quitting would “waste” them. That’s the sunk-cost fallacy.

The cleanest analogy is the movie ticket. You pay $15 for a film. Twenty minutes in, it’s dreadful. Do you stay? The $15 is gone either way — you don’t get it refunded for sitting through the rest. So the only real question is: given that the $15 is spent, how do I want to spend the next 90 minutes — bored in a dark room, or doing literally anything else? Walking out is the right call. Staying “to get your money’s worth” just adds 90 wasted minutes to the 15 wasted dollars. You can’t get your money’s worth out of money that’s already gone.

Worked example — averaging down, framed two ways

You own 100 shares of a company bought at $50, total cost $5,000. The price falls to $30. You have $3,000 of fresh cash to invest somewhere. Should you buy 100 more shares of this company at $30?

Watch how the framing changes the felt answer even though the money is identical:

The “average cost” of $40 is a number that exists only in your head. The market has never heard of it and will never reward it. If the company at $30 is genuinely the best use of your $3,000, buy it — but buy it because it’s the best forward bet, not because it drags down a break-even line that nobody but you can see. If you’d rather put the $3,000 into an index fund, then that’s your answer, and the $50 you once paid changes nothing.

The pitfall — “I don’t want to waste what I’ve put in”

The fallacy disguises itself as thrift, which is why it feels virtuous. “I’ve already put so much in, I can’t quit now” sounds responsible. It’s the opposite. Continuing to spend good resources to honour spent ones doesn’t redeem the spent ones — it just guarantees you lose more. Thrift is about future dollars. The dollars behind you are not yours to be thrifty with anymore.

When it matters

Sunk-cost reasoning shows up the instant you’re behind on something — a falling stock, a renovation over budget, a degree you’ve lost interest in, a relationship, a strategy that isn’t working. The bigger the amount already sunk, the louder the fallacy screams. Train the reflex now: the size of what you’ve already lost is a measure of past events, never a reason for future ones.

Escalation of commitment — good money after bad

If the sunk-cost fallacy is the static error — holding because of what you paid — then escalation of commitment is its moving, compounding cousin: actively pouring more resources into a failing course of action specifically to justify the resources you’ve already sunk. It’s the phrase your grandmother knew: throwing good money after bad.

The textbook real-world case is Concorde. Britain and France kept funding the supersonic jet for years after it was clear it would never make money, precisely because they’d already spent so much — abandoning it would mean admitting the prior billions were wasted. So they spent more to avoid feeling the loss. Economists sometimes call the sunk-cost fallacy the “Concorde fallacy” for exactly this reason. The extra spending didn’t rescue the earlier spending; it joined it in the hole.

In investing, escalation looks like adding to a losing position with no new thesis — buying more only because the price fell and you want to “get back to even faster.” Notice the tell: the justification points backward (your purchase price, your paper loss) rather than forward (a fresh reason the investment is now a better bet than the alternatives). Averaging down can be perfectly rational — if the lower price genuinely makes the future return more attractive and you’d buy it on the merits. Escalation is when the only reason you’re buying more is to defend a past decision.

Worked example — two investors, same purchase, opposite logic

Both bought at $50; both watch it fall to $30; both buy 100 more shares. Identical action — but only one is escalating.

Same trade, completely different decision quality. Investor B has let a backward-looking goal — don’t make the first loss real — drive fresh capital into a bet they never re-evaluated. That’s escalation, and it’s how a small mistake grows into a large one.

The disposition effect — selling winners, riding losers

Here’s where the psychology gets specific and measurable. The disposition effect is the documented tendency of investors to sell winning investments too early and hold losing ones too long — to “dispose of” winners while clinging to losers. It was named by Shefrin and Statman (1985) and nailed down with real brokerage data by Terrance Odean (1998).

Why it happens — it’s prospect theory, applied

This isn’t a random quirk; it falls straight out of the Prospect Theory you met in the previous lesson. Three of its pillars combine to produce the disposition effect:

• Reference dependence. You judge an investment as a gain or a loss relative to a reference point — almost always your purchase price — instead of as a final wealth level. The stock at $30 isn’t “an asset worth $30”; it’s “the thing I’m down $20 on.”
• Loss aversion. Losses hurt roughly twice as much as equal gains feel good. So realising a loss (selling at $30) means actually feeling that doubled pain, while holding keeps the loss “on paper,” where it stings less.
• Diminishing sensitivity — the convex (risk-seeking) shape of the value function in the loss region. When you’re already down, you’ll gladly gamble for the chance to claw back to even rather than accept a sure loss. So you hold the loser, hoping. On the winning side the function is concave (risk-averse), so you grab the sure gain by selling early rather than risk giving it back.

Put together: gains make you want to lock in (sell winners); losses make you want to gamble for recovery (hold losers). The reference point is the purchase price; the asymmetry is loss aversion; the gamble-when-losing is diminishing sensitivity. The disposition effect is prospect theory wearing a brokerage account.

The evidence — Odean’s PGR ≈ 1.5× PLR

Odean (1998) studied 10,000 accounts at a large US discount broker and measured two rates:

• PGR — the Proportion of Gains Realised: of all the winning positions an investor could have sold on a given day, what fraction did they actually sell?
• PLR — the Proportion of Losses Realised: of all the losing positions they could have sold, what fraction did they actually sell?

If investors were neutral about gains versus losses, PGR and PLR would be roughly equal. They weren’t. Odean found:

$\text{PGR} \approx 1.5 \times \text{PLR}$

Investors cashed in their winners at about 1.5 times the rate they cashed in their losers. The reluctance to sell a loser was systematic, large, and visible across thousands of accounts. (The effect softened in December — when tax-loss selling gave people a reason to finally realise losses — which is itself a clue that the rest of the year’s behaviour was psychology, not strategy.)

The double mistake — it costs you twice

Here’s the part that turns a curiosity into a real wealth-killer. The disposition effect is wrong in two independent ways at once:

1. It’s tax-inefficient. In a taxable account, selling a winner triggers a tax bill now on the gain, while selling a loser can often be used to reduce your taxes. The disposition effect does exactly the reverse of the tax-smart move: it realises gains (taxed) early and defers losses (the useful deduction) late. You’re literally paying tax sooner than you needed to.
2. It picks the wrong horses. You’d hope at least the stock-picking was sound — that the winners you sold were about to fizzle and the losers you kept were about to recover. Odean found the opposite: on average, the winners investors sold went on to outperform the losers they held onto. So they sold the stocks that kept rising and kept the stocks that kept lagging. The behaviour didn’t just cost taxes — it actively lowered returns on top.

So it’s not “a harmless preference.” It’s a tax mistake and a returns mistake, stacked. Tap through the holdings below, make your own sell/hold calls, then reveal the pattern and the data.

Sunk cost or genuinely new information?

Now the hard part — the part that separates discipline from dogma. “Ignore sunk costs” does not mean “ignore that the price fell.” A price drop can be two completely different things, and your whole job is to tell them apart:

• A sunk cost / loss-aversion trap: the price fell, nothing about the business changed, and you’re tempted to hold (or buy more) purely to avoid realising the loss or to “get back to even.”
• Genuinely new information: the price fell because something real changed — earnings collapsed, a key product failed, the regulator opened an investigation, the thesis broke. That’s not a sunk cost. That’s the world telling you the forward outlook is worse, and ignoring it would be its own mistake.

The drop itself doesn’t tell you which one you’re facing. A stock down 40% on a fraud revelation and a stock down 40% on a market-wide panic look identical on the chart. What separates them is why, and whether that why changes the future.

The clean-slate test

There’s one question that cuts straight through, and you should burn it into memory:

“Knowing only what I know now — not what I paid, not what I’ve lost — would I buy this investment today?”

This works because it surgically removes everything backward-looking (purchase price, paper loss, ego, “break-even”) and leaves only the forward decision. The answers map cleanly:

• Yes, I’d buy it today → hold it (and maybe even buy more — on the merits, not to average down).
• No, I wouldn’t buy it today → you shouldn’t be holding it either, because holding is the same decision as buying. Holding a stock is choosing it over everything else you could own with that money, every single day.

That last line is the quiet killer: holding is a decision you re-make every day. Refusing to sell a loser isn’t “doing nothing” — it’s actively choosing, every morning, to keep your money in something you might not buy fresh. The clean-slate test forces you to see the hold as the active choice it really is.

Worked example — same 40% drop, opposite verdicts

Same percentage loss, same temptation to “wait for break-even” — but the clean-slate test gives opposite, correct answers, because it asks about the future, which is the only thing that differs between the two.

Sort these statements into what to ignore versus what to act on:

Putting it together

Two stacked traps, one antidote. Sunk costs whisper that the past should steer the future; escalation turns that whisper into fresh money in a hole; the disposition effect is the specific, measured form it takes in a portfolio — selling winners, riding losers, paying twice. The clean-slate test is the lever that pops all three at once.

A mixed recap — it pulls from everything above:

Key Takeaways